Two people left Google within the same stretch of 2026, both told their stories to the same reporter, and the stories contradict each other in a way that matters more than either one does on its own.
Yousuf Imran, a 41-year-old account executive at Google, earned close to a million dollars in 2025 and left that April to found Mangosteen Studio, an AI product lab building tools for sales teams. Candice Bryant, an internal communications manager at Google who spent the previous 16 years at the CIA (the US intelligence agency), left in October 2025 to build Cello, an app that sends users daily prompts for practical AI use. Jacob Zinkula, the Business Insider reporter who wrote both as-told-to essays, ran Imran’s as part of a June 27 roundup of six Google departures.
Same employer. Same reporter. An eighteen-month window. Read the two accounts side by side and you get two incompatible explanations for why capable, well-paid people are walking out of one of the best jobs in tech.
Kill the obvious misread first. This is not a story about Google struggling. Google’s cloud revenue grew 63% past $20 billion for the first time in the same period, with its backlog nearly doubling to over $460 billion, even as the company kept cutting headcount. Nobody here is fleeing decline. They’re fleeing what success currently looks like from the inside.
Imran’s reasoning is about equity, and it’s worth being precise about what that word is doing. A salary is a fixed claim on a company’s revenue, paid whether the company’s value doubles or stays flat. Equity is a variable claim on the value itself. Google can pay someone close to a million dollars a year and they still own essentially none of the outcome they’re helping produce. A founder, or an early hire at a frontier lab, owns a direct slice of that outcome. That’s the whole gap. Imran called it “AI FOMO.” It means a Google salary, however large, doesn’t compare to what someone can capture by holding equity in the AI boom instead of drawing a paycheck from it.
That gap isn’t closing. It’s widening at the top and staying flat everywhere else. Jensen Huang, the CEO of Nvidia (the chipmaker whose processors run most frontier AI models), personally reviews and signs off on pay for all 42,000 Nvidia employees and says he raises the company’s compensation spend every time. His philosophy is blunt: “I pay my employees as much as I can.” That approach has already made several people on his management team billionaires. Meanwhile Nvidia’s own securities filings put its median employee compensation at roughly $301,000, a rounding error next to what those billionaires hold in stock. The pattern repeats up the supply chain. Samsung’s semiconductor division just ratified a deal giving workers bonuses averaging roughly $340,000, with memory-division employees seeing closer to $400,000 in a single year. TSMC, the Taiwanese company that manufactures Nvidia’s chips, told staff at a companywide town hall their profit-sharing payouts would rise more than 30% this year.
Those are compensation numbers, not equity numbers, and the difference is the whole point. Compensation is what a company decides to pay you out of profits it already made. Equity is what you hold regardless of what the company decides. Even Nvidia’s pay philosophy, the most generous in the industry, doesn’t hand a rank-and-file engineer what Imran left Google to chase: ownership in whatever gets built next. Big Tech and its suppliers can raise salaries and bonuses as high as they want. The gap between employee and owner in an AI boom doesn’t close. It just gets a higher floor. Imran gave up the floor to bet on the ceiling.
That race is real, and the ground under it is shakier than the headlines suggest. US tech companies announced 85,411 job cuts in just the first four months of 2026, a 33% jump over the same period in 2025. But the story of who’s actually causing those cuts is contested. Marc Andreessen, the venture capitalist who cofounded the VC firm a16z, has argued the layoffs mostly trace back to pandemic-era overhiring and higher interest rates, calling AI a convenient “silver bullet excuse.” Oxford Economics, an economic research firm, concluded that firms “don’t appear to be replacing workers with AI on a significant scale.” Whether the cuts are genuine AI restructuring or convenient cover for cuts companies wanted to make anyway, the effect on someone deciding whether to stay is identical: the ground looks unstable, and betting on yourself starts to look like the safer move.
Bryant’s reasoning has nothing to do with equity. She watched Google’s own AI tools go live internally and noticed that almost no one outside the building was using them. Her read: the real divide in AI right now isn’t between Silicon Valley and policymakers. It’s between Silicon Valley and everyone else. A technology nobody understands doesn’t transform anything, no matter how capable it is.
The data backs her up more than the “everyone’s already using AI” narrative wants to admit. Stanford’s Institute for Human-Centered AI, in its 2026 AI Index, found a 50-point gap between AI experts and the general public on whether the technology will help their jobs, 73% versus 23%, with a similar gap on the economy. McKinsey’s most recent global AI survey found 78% of organizations now use AI in at least one business function — 71% specifically use generative AI regularly — but only 39% can point to any actual enterprise-level earnings impact from it. Even at the consumer level, only about 3% of American households were paying for an AI subscription as of February, and ChatGPT converts just 5% of its weekly users into subscribers. People are touching AI constantly. Understanding what it’s worth to them, or trusting what it does to their job, is a different and much less solved problem. Usage climbed. Comprehension didn’t climb with it. That’s Bryant’s whole thesis, and the numbers say the gap is wider than the adoption headlines let on.
Here’s the part that should bother you more than either individual exit. Only one of these gaps is currently being funded. Every dollar chasing the AI moment, whether it’s Imran’s own startup or Nvidia’s supply chain sharing record profits, is chasing the equity and compensation gap. Almost none of it is chasing the comprehension gap. Nobody has built a venture-scale business around closing the distance between what AI can do and what an ordinary person believes it can do, because that business doesn’t monetize as fast as another sales tool or a bigger bonus pool.
That’s the aspiration logic running backward. The industry keeps holding up the equity-chaser as the person to aim for, slightly ahead, in the direction everyone’s told to want. Meanwhile the actual scarcity, a trust gap that Stanford is measuring at 50 points and rising, sits mostly untouched, because it doesn’t pitch as cleanly on a term sheet.
Two people left the same building for opposite reasons. One is betting the AI moment pays off in equity. The other is betting it pays off in understanding. The market has already priced its favorite. If everyone’s racing after the equity gap, what would it take to go after the one nobody’s pricing yet?
Sources
Yousuf Imran, as-told-to Jacob Zinkula, Business Insider (via Yahoo Finance), June 28, 2026: https://finance.yahoo.com/technology/ai/articles/left-google-making-nearly-1m-111556351.html
Candice Bryant, as-told-to Jacob Zinkula, Business Insider, July 2026: https://www.businessinsider.com/cia-google-employee-quit-ai-tech-business-entrepreneur-career-advice-2026-7
Alphabet Inc., Q1 2026 Earnings Release (SEC 8-K): https://www.sec.gov/Archives/edgar/data/0001652044/000165204426000043/googexhibit991q12026.htm
“How Nvidia CEO Jensen Huang treats employees of the world’s most valuable company,” Fast Company: https://www.fastcompany.com/91552520/how-nvidia-ceo-jensen-huang-treats-employees-of-the-worlds-most-valuable-company
“Samsung chip workers vote to accept $340,000 average bonus…,” Tom’s Hardware: https://www.tomshardware.com/tech-industry/samsung-chip-workers-vote-to-accept-340000-average-bonus-ending-months-long-strike-threat
“TSMC CEO Pledges 30%-Plus Incentive Bump as AI Profits Soar,” Bloomberg: https://www.bloomberg.com/news/articles/2026-05-27/tsmc-s-ceo-pledges-30-plus-incentive-bump-while-ai-profits-soar
“Tech layoffs climb as AI remains top driver,” CFO Dive: https://www.cfodive.com/news/tech-layoffs-rise-ai-remains-top-driver-us-workforce-cuts/819640/
“Marc Andreessen says AI layoffs are a farce…,” Fortune: https://fortune.com/2026/03/31/marc-andreessen-ai-layoffs-silver-bullet-excuse-overhiring/
“Evidence of an AI-driven shakeup of job markets is patchy,” Oxford Economics: https://www.oxfordeconomics.com/resource/evidence-of-an-ai-driven-shakeup-of-job-markets-is-patchy/
Stanford HAI, 2026 AI Index Report, Public Opinion: https://hai.stanford.edu/ai-index/2026-ai-index-report/public-opinion
“The State of AI: Global Survey,” McKinsey: https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai
“A handful of Americans pay for AI. Will you have to?” NPR: https://www.npr.org/2026/06/04/nx-s1-5791661/chatgpt-gemini-claude-subscription-revenue-openai
